The aim of this thesis is to develop machine learning based model for the heat transfer prediction of a Thermosiphon system.
Passive, self-sustaining heat removal systems can be applied in spent fuel pools as a redundant, inherently safe heat removal system. Within the scope of this research project, experimental investigations are performed to characterize the heat transfer behavior of long heat pipes and thermosiphons and a comprehensive experimental database is generated. As an innovative approach, we plan to develop and extend machine learning based alogorithms, which will be purely data-driven. The work will mainly emphasize on the application of deep neural network (DNN) using the TensorFlow library. It is also planned to do a comparison between commonly used correlations and DNN.
- Basic knowledge thermo-fluid dynamics.
- Nice to have: Basic knowledge of Pyhton, machine learning and data processing.
Start: As soon as possible
More details are available here.